relay coil and load snubbing

Should the diode be connect across the relay load or the
relay coil?

Thanks,

    - Matt

There are two issues here - snubbing the back EMF across the relay coil and 
snubbing a possibly inductive load across the relay contacts.

The issues are similar. Voltage across an inductor is proportional to the 
time rate of change of the current through it. V = L di/dt so if you try to 
stop the current in zero time (infinite di/dt), the V will also approach 
infinity. This is why you must clamp the voltage. In practice, di/dt is 
always less than infinite but V can easily be 2-4 times or more the value 
of the steady state DC voltage across the coil. So a 12 VDC coil can 
generate 48 volts or more of kickback when switched from the driven to 
undriven state.

Clamp the relay coil like this:
http://www.systronix.com/expansion/8x1wire/DS2406_relay_drive.pdf

Then if the relay *contacts* are driving an inductive load you have the 
same problem on that end, and the contacts will arc and pit as they try to 
break an inductive circuit (you don't have the same problem when they close 
to drive in inductive load, just when they break an inductive circuit with 
current already flowing). You can't generally put a diode across all loads 
(especially AC ones!) so you use an R-C snubber like this:
http://www.cornell-dubilier.com/mpdf/quench.pdf

Here are some TRIAC snubber circuits, and the same principle holds for 
mechanical relays:
http://www.epanorama.net/circuits/semiconductor_relays.html#snubber

Finally, a caution: if you are switching mains or very high current loads, 
and you aren't an EE or otherwise qualified, it's best to consult an EE or 
other qualified person to avoid frying something (your fingers or your 
house, for example). You must use appropriate components for such things 
(UL/CS/VDE/etc), in an appropriate circuit and there are a host of related 
safety guidelines. The consequences of a mistake could be significant.

Regards

Bruce


_______________________________________________
TINI mailing list
TINI-6tN4nzCoH/[email protected]
To UNSUBSCRIBE, edit your profile, or see list archives:
http://lists.dalsemi.com/mailman/listinfo/tini